Network governance. A comparison of four Dutch regions in their approac towards the housing of EU migrant workers

(1)

Network Governance

A Comparison of Four Dutch Regions in

Their Approach Towards the Housing of EU

Migrant Workers

Marc Kuppens

Supervisor: Jolien Grandia

Department of Public Administration Faculty of Management Sciences

Radboud University Nijmegen October 2016

(2)

Preface

Before you lies my masterthesis “Network Governance: A Comparison of Four Dutch Regions in Their Approach Towards the Housing of EU Migrant Workers”. This

masterthesis will be my final project for the master Comparative Public Administration at the Radboud University Nijmegen. Furthermore, will my live as a student end, now I finished my thesis.

By interviewing respondents from municipalities, labor associations and housing

associations, I tried to analyse characteristics of a successful governance network. The research process and its outcomes can be found in this thesis.

Writing this thesis was a long process. A process with a lot of obstacles, but in the end I am very satisfied with the overall research process and its outcomes. I would not have succeeded without the help of my supervisor Jolien Grandia, who guided me during the research process towards the proper direction. Furthermore, do I want to thank my girlfriend Isa and my family for their support.

I hope you enjoy reading.

Marc Kuppens

(3)

1. Introduction

After over half a decade of intense war on the European continent between the years 1939 and 1945, the prevailing idea in Europe was that situations such as World War II should never be repeated. To ensure this, the foundations where laid for a

cooperation between states that was never seen in Europe before. In the first stages, this cooperation aimed at pooling the necessary resources for war: steel and coal. For this reason the European Coal and Steel Community was formed. The European Coal and Steel Community has evolved in the last century into what we now know as the European Union. One of the cornerstones of the European Union is the border-free movement of citizens between most of the member states, as agreed on in the Schengen Agreement of June 18th_{1985. The right to freely cross the border between}

member countries of the European Union can be seen as positive and progressive. It involves many positive effects, such as the fact that the free movement of citizens of the European Union has had a large positive impact on the economy, since Eastern European migrant workers are in general cheap labour forces (Passport to trade 2.0, 2015). Unfortunately, the right to move freely from country to country and to work throughout Europe has some consequences that are less positive.

In the year 2004, a large group of Central and Eastern European countries became members of the European Union1_{(EU). Their membership of the EU opened the}

border towards Western Europe for large groups of migrant workers from these Central and Eastern European countries. Since the Western European states in general have stronger economies and a relatively stable labour market with higher wages, a lot of Central and Eastern European citizens used the possibility of crossing the intra-European borders freely to work in these countries (Holtslag, Kremer and Schrijvers, 2012: 7). This flow of migrants towards Western European countries grew rapidly over the last ten years. Therefore the number of Eastern European migrant workers living in Western European countries has also been growing (Migration policy, 2011). The growing number of EU migrant workers results in housing

problems in the host countries. Also in The Netherlands, one of the countries where many migrant workers migrate to, there is a lack of proper housing for these EU migrant workers. In order to deal with the housing problems of EU migrant workers, the Dutch government appointed nine ‘priority regions’ in which municipalities, housing associations and labour associations have to work together in order to provide more and better housing for EU migrant workers.

The problem of finding enough and suitable accommodation for Eastern European migrant workers is a complex problem, since different actors are involved, such as municipalities, labour associations and housing associations. As a result, any sole actor cannot solve the problem. The resulting need for cooperation is linked to the different resources that different actors possess. A combination of resources is

(6)

needed to tackle the problems surrounding the housing of EU migrant workers, therefore the actors governing these different resources need to work together. (“Nationale verklaring van partijen betrokken bij de (tijdelijke) huisvesting van EU-arbeidsmigranten”, 2012).

1.1. Comparison of four Dutch regions

The problems surrounding the realization of enough and suitable housing for EU migrant workers are present in many regions in The Netherlands.

Flexwonenarbeidsmigranten (2016), a governmental research center for housing of EU migrant workers, appointed nine ‘priority regions’ within the Netherlands. In these regions, the need for provision of more and better housing for EU migrant workers was most urgent, since most EU migrant workers lived in these regions and an extensive shortage of proper housing existed. In this research the focus will be on four of these nine regions: Holland Rijnland (positioned around the Dutch city of Leiden in the province of Zuid-Holland), West-Friesland (a region north of Amsterdam in the province of Noord-Holland), Metropoolregio Eindhoven (the region around the Dutch city of Eindhoven in the province of Noord-Brabant) and Noord-Veluwe (a region geographically situated in the middle of The Netherlands in the province of Gelderland). In these regions, there is a large presence of EU migrant workers and an urgent demand for suitable accommodations for these migrants. More on the case selection will follow in the section concerned with methodology.

In each of the four regions in this research, the involved actors (municipalities, labour associations and housing associations) signed a housing agreement for EU migrant workers. These housing agreements function as a tool for regulating the cooperation of different public- and private actors that are involved in the process of finding more and better housing for EU migrant workers. This cooperation of different actors is a typical example of network governance; since different actors, with different interests, have to work together in order to tackle a joint problem. More information on the housing agreements that were signed in the four regions can be found in the next chapter of this research: The case description.

1.2. Problem definition, central question and sub-questions

This research focuses on the compliance of municipalities, housing associations and labour associations with the regional agreements on the housing of EU migrant workers in four Dutch regions. Special attention will be devoted to the cooperation of the involved actors in providing more and better housing for EU migrant workers. This will all be placed in the context of the refugee crisis. The goal of this research is to gather insight in if, and how, governance networks operate and if they result in a proper output (in this case successful realization of enough EU migrant workers housing). The four Dutch regions and the governance networks within these regions will function as cases in this research. In order to fulfil the goal of this research, interviews with actors within the four regional governance networks will be conducted.

(7)

The four regional governance networks will be compared for certain specific

characteristics that function as indicators of governance network success: deemed necessity, goal consensus, active actors in the network, trust and steering. These characteristics are acknowledged by several authors on governance network as the most important characteristics for governance network success (Provan and Kenis (2008), Klijn, Edelenbos and Steijn (2010), and Torfing and Sorensen (2014)). These characteristics of governance networks will from now on be called governance

network characteristics. In order to get an overview of the governance network characteristics in each of the four regions, all respondents were interviewed about the network characteristics of their regional governance network. Furthermore, all respondents were asked about the influence of the refugee crisis on the compliance with the regional housing agreements, since it is expected that the refugee crisis has a negative influence on the efforts on the housing of EU migrant workers.

Concluding, the following question will be the central question of this research:

How do the governance network characteristics influence compliance of actors within the four regional governance networks with the regional agreements on the housing of Eastern European migrant workers in the context of the refugee crisis?

Several sub-questions should be answered before the central question of this research can be answered:

- To what degree do the governance network characteristics influence the compliance of the actors within the four regional governance networks with the regional housing agreements?

- To what degree there is compliance of the actors within the four regional governance networks with the regional housing agreements?

- To what degree does the refugee crisis, as a context factor, influence the compliance of actors within the four regional governance networks with the regional housing agreements?

This research will both contribute to theoretical (public administrational) as well as societal knowledge. It contributes to the theoretical knowledge of network

governance, by evaluating and comparing different theories on governance networks and specific characteristics of these networks. By doing this, a theoretical foundation of combined notable works on governance networks is formed, that functions as foundation for this research. This theoretical foundation will encompass the most important network characteristics for governance network success. By researching these governance network characteristics for the four governance networks in this research, knowledge is gathered about the effects of these governance network characteristics on the success of the governance networks. This will all be conducted on a regional level, which is a governance level that not many researchers focus on. Therefore, it is interesting to investigate this layer of governance. Finally, there is

(8)

aimed for answering the question about which combination of network characteristics will result in governance network success (in the case of the proper housing of EU migrant workers). This will result in a better understanding of the working of

governance networks and the specific characteristics that they can possess.

On the societal angle, this research will be a first step in the direction of determining the best way to ensure enough and suitable accommodation for EU migrant workers. By comparing the four regional governance networks in their compliance with the agreements on the housing of EU migrant workers, and by researching the determents for (non-)compliance, this research can inform policy makers about several ways to cope with problems surrounding the housing of EU migrant workers. This research will therefore be a policy document that can be used as

recommendation on the issue of housing EU migrant workers. It can also help to identify pitfalls of the four regional governance networks.

1.3. Readers guide

Governance networks have some specific network characteristics. The influence of these governance network characteristics on the success of the four regional governance networks will be the central point of interest in this research.

Furthermore, the influence of the refugee crisis on compliance with the regional agreements on the housing of EU migrant workers will be researched.

In order to answer the central question of this research, the structure of this paper will be as follows: in the second part of this research, a case description will follow, in which the context of the housing problem will be discussed. The third part of this research will function as a theoretical foundation, in which different notable works on governance networks will be coupled. Subsequently, in the fourth section of this research, the methodological considerations of the research will be explained as well as the case selection. Then in the fifth part, the analysis of the data will follow.

Finally, in the sixth section of this research, conclusions will be drawn on the findings of this research, the research will be reflected on and recommendations for further research will be given.

(9)

2. Case description

This chapter will function as a case description for this research. The goal of this chapter is to describe the research problem in detail in order to be able to place it in a broader context (the context of the refugee crisis). This is done by describing the housing problem in detail and by connecting the housing problem to the different groups of EU migrant workers that need accommodation. In addition, the origins of the refugee crisis will be explained in order to make a sketch of the broader problem that is faced.

In The Netherlands, the country on which this research will be focused, the number of Polish migrants quadrupled from 35.542 Polish migrants in 2004 to over 150.000 in 2016 (NOS, 2016). This rapid growth in the number of EU migrant workers causes housing problems, since all these migrants need a place to stay in the host-country. The main problem, in the case of the Netherlands, is that there already is a shortage of low-income housing for a couple of decades (Algemeen Dagblad, 2015). The flow of migrants from Eastern European countries, who also need low-income housing, enlarges these problems. The fact is, that EU migrant workers can be placed in different groups, and the type of EU migrant worker has consequences concerning the need for realizing housing for the immigrant. In general, a division can be made between short-stay, mid-stay and long-stay. This division is present in the policy approaches in all four Dutch regions, but is best described by the regional agreement of the Metropoolregio Eindhoven. De Tweede Regionale Verklaring

Arbeidsmigranten in Zuidoost-Brabant Versie 27-11-2013 states the following:

“One can speak of short-stay of EU migrant workers when they intend not to stay longer in the host-country than four months. Mid-stay is when an EU migrant worker wants to stay longer than four months in the host-country, but intends to leave again. Long-stay is when the EU migrant worker wants to settle for a longer period in the host-country (most of the time for several years). The division between mid-stay and long-stay is not always clear, whereas in some regions there is only a

division between short-stay and long-stay”. (De Tweede Regionale Verklaring Arbeidsmigranten in Zuidoost-Brabant Versie 27-11-2013)

Policy differs when it comes to housing of EU migrant workers with different times of stay. In general, the regional governance networks and agreements are focussed on short-stay and mid-stay migrants. For these groups suitable housing should be realized. The EU migrant workers that intent to stay longer in the country should also make use of this housing in the first stages of their stay. Furthermore should they, in a certain amount of time, flow towards the regular housing processes at housing associations, like other Dutch people. The scope of this research is limited to

(10)

short-stay and mid-short-stay, since this type of accommodation is affected most by the refugee crisis and therefore most interesting to conduct research upon.

With the need for realization of more and better housing for EU migrant workers, the involved actors are charged with a big task. Accommodation is a vital necessity and the first requirement to ensure a living in the host-country. In most cases, Eastern European migrant workers do not have a lot of money available for accommodation. As such, it is critical to provide these migrants with proper low-budget housing. It is also important to keep in mind that not all EU migrant workers ask for the same type of accommodation as elaborated upon earlier in this chapter. They are in general, and also in this research, divided into three groups: short-stay, mid-stay and long-stay (Platform 31, 2015).The task of realizing enough and suitable housing for the EU migrant workers in The Netherlands is hard, since different actors have to work together. A thing that makes the task even more complicated is that there is no reliable information on the number of EU migrant workers that is present in the regions. Therefore, in all regions, the agreements were accompanied by quick scans investigating the number of migrant workers present in the regions. Because of the lack of information on the number of migrant workers, there is almost no insight in how many migrants need a place to stay and what their housing needs are. In order to detect who enters the country and needs a place to stay, the different actors within a governance network should first map the scope of the problem. This was the goal of quick scans that were conducted in the regions. These quick scans mapped the number of EU migrant workers in each region, by looking at existing governmental documentations.

The search for accommodation is a multi-agent undertaking, which means that different actors will have to cooperate in order to tackle the problems. Furthermore, cooperation is needed between governmental actors (municipalities) and

non-governmental actors (housing associations and labour associations), who both play a vital role in the search for enough and suitable accommodation. A good organization of this network of governmental and non-governmental organizations is key to the effective tackling of the housing problems for Eastern European migrant workers. In all four regions included in this research, a similar approach towards the problems can be noticed in which different private- and governmental partners work together in finding solutions for the problems surrounding the accommodation-mission for EU migrant workers. Following an initiative from the Dutch Ministry of Home Affairs, an agreement was signed in all four regions between the different parties involved: municipalities, housing associations and labour associations. In these agreements, the different actors, with different interests, agreed on finding joint solutions to the problems that they were facing. This cooperation is a typical example of network governance, in which actors from different branches, with different interests, combine their knowledge and resources in order to tackle a problem that affects all actors. Many similarities can be noticed between the agreements in the four regions. There is a general trend in what was asked of the three main branches in each region. The

(11)

municipalities were asked to ensure enough governmental capacity to come up with solutions for the housing problem and to mediate between the different actors involved in the process of solving the housing problem. The municipalities have a facilitating role in the agreements, they should in short, facilitate the initiatives on the housing of EU migrant workers. The labour associations were asked to report on the number of Eastern European migrant workers working for them in order to establish the size of the problem in that region. Furthermore, they will have to cooperate with municipalities to ensure housing of a certain minimum standard. The housing

associations were asked to cooperate in finding enough suitable accommodation for the migrant workers. Theoretically, the housing problem can be combatted effectively when all three branches work together and comply with their regional agreement. It is expected that the four regional governance networks differ in the success of realizing more housing for EU migrant workers, since it is likely that they at least differ on some of the governance network characteristics.

2.1. The refugee crisis

The regional agreements on the housing of EU migrant workers were signed in 2013 and 2014 by the involved actors. At that time, there was an ongoing war in Syria. The war in Syria began with the events of the Arab Spring. During the Arab Spring, which began in the year 2011, mass protests took place in several North African and Arab countries. The protests were aimed at the dictators that ruled most of the countries in the region. The protests were characterized by a call for democracy and civilians saw their chance to attain more control over their own lives (BBC, 2016). Syria was one of the countries involved in this Arab Spring. The president of Syria, Bashar al-Assad, ruled the country in a heavy-handed manner. It could even be said that he was a dictator, and to a large portion of the population of Syria there was no doubt that he was a dictator (News.com.au, 2015). The mass protests in Syria soon escalated into a civil war between many rebel groups. The Syrian Assad government even used chemical weapons on its own citizens in order to free the country from, whom they were considered to be, terrorists (BBC, 2016). The ongoing civil war in Syria has resulted in a mass flow of Syrian refugees towards the neighbouring countries. Lebanon was the main country of refuge, but also Saudi Arabia and Turkey received many refugees (BBC, 2016).

The civil war took a completely new turn when Islamic State of Iraq and the Levant (from now on referred to as IS) conquered large portions of Syria and Iraq and proclaimed itself to be the new caliphate (an Islamic state-type believed to be ruled by a successor of the Prophet Muhammad) (BBC, 2016). This caliphate was ruled by Caliph Abu Bakr al-Baghdadi. Large numbers of Syrians once again fled to

neighbouring countries as a result of the cruelty of IS towards everybody that was, in their eyes, not a good Muslim (BBC, 2016). The civil war in Syria is still ongoing nowadays and many Syrian refugees have lost their hope on a swift resolution of the conflict. As a result, they try to flee to European countries, mainly via Greece and Italy, where they want to build a new life for themselves. Because of this

(12)

development, Europe has seen more than half a million refugees cross its borders in the last two years (NOS, 2015). Refugees tend to flee to Europe, instead of Syria's neighbours, with the hope of better (wealthier) lives, since they lost their hope for a quick resolution to the conflict.

The continuous stream of refugees from non-European countries has its consequences for the intra-European migrants and their chances of finding accommodation. A number of refugees receives a legal status and will stay in the host-country for a prolonged period. When looking for suitable accommodation the Eastern European migrant workers and refugees have to compete about suitable housing. The movement of Syrian refugees towards Europe is regarded as a context factor in this thesis, which influences the search for more and suitable

accommodation for Eastern European migrant workers. It can have consequences on the regional level and the agreements made by the governance networks in the four Dutch regions included in this research.

The goal of this chapter was to explain the research problem and the research case more detailed. The problem of finding proper housing for EU migrant workers is a complex problem because of several reasons. This complexity is in the first place present due to the different types of EU migrant workers that need a place to stay. This research will be focussed on two of these types, short-stay and mid-stay. Furthermore, the need for cooperation of different private and public actors adds to the complexity of solving the housing problem. In addition, is it important to keep the broader context of the refugee crisis in mind in the process of problem solving, since it is expected that the housing of refugees has negative influence on the housing of EU migrant workers. In the next chapter the theoretical foundation of governance networks, which is used for this research, will be build.

(13)

3. Theoretical framework

In this chapter, the theoretical foundation for this research will be established. The focus of this research will be on governance networks and specifically the

compliance of the governance networks in four Dutch regions with the agreements on the housing of EU migrant workers. Compliance by the governance networks with these agreements is seen as success for the governance networks. When this is the case the goal of the governance network is reached, and the network can be seen as successful. In order to say something about the working of the four Dutch

governance networks later on, there will now be drawn a model of a successful governance network. In order to do this, there will first be explained what a governance network encompasses. Furthermore, there will be explained why governance networks are wished for in solving certain societal problems. Then, a section will follow about how governance networks come about. When these more general notions on governance networks are explained, the chapter will move further into the specific characteristics of a successful governance network. In the end, everything will be summarized in a drawn model of a successful governance network.

3.1. What is a governance network?

The central theoretical objects in this research are governance networks. Torfing (2012: 101) defines a governance network as follows:

“Horizontal articulations of interdependent, but operationally autonomous, actors from the public and/or private sector who interact with one another through ongoing negotiations that take place within a regulative, normative, cognitive, and imaginary framework; facilitate self-regulation in the shadow of hierarchy; and contribute to the production of public regulation in the broad sense of the term”. (Torfing, 2012: 101)

As mentioned above, the search for more and better accommodation for Eastern European migrant workers does involve different actors. The three most important branches in the problem solving of the housing of EU migrant workers are

municipalities, labour associations, and housing associations. These involved actors form a governance network, since they are mutually dependent on each other and need a combination of resources that they possess in order to solve the problems surrounding the housing of EU migrant workers (Rhodes, 2007). Governance networks are an utterance of governance, in which governance means:

“Lateral and interinstitutional relations in administration in the context of the decline of sovereignty, the decreasing importance of jurisdictional borders, and a general institutional fragmentation. Where in the administrative state is less bureaucratic, less

(14)

hier-archical, and less reliant on central authority to mandate action”. (Frederickson et al., 2015: 236)

The meaning of governance, according to Levi-Faur (2012), is quite disputed. The term can have many meanings and is therefore used in different ways (Peters, 2012). Obviously therefore, the given definition of governance is not the only good definition of the subject. The use of governance as a vehicle for execution of a policy won ground in the 1990s due to a growing complexity of social and political life. When we compare our present society with the one a hundred years ago, complexity of society has grown rapidly since the 1990s (Klijn & Koppenjan, 2012: 598). The problems that come forth out of society have grown in complexity as well. Complexity of social and political life has grown mainly due to two factors as elaborated on by Torfing (2012: 100): 1) Globalization and 2) new ideas on how to govern society. Globalization is a very clear-cut cause for the growing complexity of our present society (Jordan, Wurzel & Zito, 2005: 479). Take, for example, the new European governmental layer that is currently of major importance and which affects many actors on national and semi-national levels. The new thoughts on how to govern society, on the other hand, ask for more elaboration. As Torfing (2012) stated about current society, different actors have to work together in order to come up with solutions for complex

problems. This has to do with the abandoning of hierarchical forms of government and the partial replacement of these forms by more horizontal forms of governance. In the latter, only a combination of resources possessed by the different actors in the undertaking of governance can solve nowadays complex problems.

According to the definition of governance networks used in this research, we speak of a governance network when the link between the involved actors is a horizontal one and the actors are dependent on each other. This applies to the cases in this research, since there is no hierarchical structure present in the regional governance networks and the actors need to cooperate in order to solve the housing problem. Furthermore, the actors within the network are a mix of public as well as private actors, public actors in the case of municipalities and private actors in the case of labour associations and housing associations. Additionally, the governance network should be engaged in a regulative debate in the shadow of hierarchy. This means that the network is self-regulatory, but that the network should operate within a hierarchical context, which is also the case for the governance networks in this research. The goal of the network should be to contribute to the solving of problems in the public sphere, which is the case for solving the housing problem for EU migrant workers. It can be seen that these characteristics are present in the cases in this research and therefore theories on governance networks are used as theoretical foundation for this research.

3.2. How do governance networks come about?

The European debate on governance networks began in the 1990s, when the

prevailing idea among theorists was that traditional forms of government were unable to solve complex, multidimensional problems (Torfing and Sorensen, 2014).

(15)

Moreover, according to Torfing and Sorensen (2014), we may not interpret the emergence of governance networks as a shift from government to governance, since there is no such thing as a zero-sum game between government and governance. This is to say that when governance emerges, government vanishes. Governance networks can help the traditional government in reaching goals outside of their traditional domains.

Provan and Kenis (2008) state different benefits of network governance in comparison with the historically hierarchical structure of government, such as: “enhanced learning, more efficient use of resources, increased capacity to plan for and address complex problems, greater competitiveness, and better services for clients and customers” (Provan and Kenis, 2008: 1).

There are several reasons to establish a governance network in order to solve problems in the public sphere. These reasons mostly have to do with the interdependency of the actors, grown complexity of problems, functionalist approaches and legitimacy.

Klijn en Koppenjan (2012: 591) state that many authors writing on the subject of governance networks agree on the interdependence between the actors within the network as one of the main reasons to form a governance network (e.g. Scharpf, 1978; Marin and Mayntz, 1991; Rhodes, 1997; Agranoff and McGuire, 2003; Klijn and Koppenjan, 2004). Because of the interdependency, cooperation is needed and the governance network can facilitate this cooperation. Beside the argument of interdependency for the initiation of governance networks, the grown complexity of societal problems over time is another issue that adds to the value of governance networks for problem solving. This growing complexity asks for new forms of regulating societal problems and, therefore, governance networks are a logical functional response to solve the more complex problems that emerge (Torfing, 2012: 104). In this view, governance networks can be seen as a new turn in

governmentality. Since markets and hierarchy are both not the best way of governing society, a new trend in governance has emerged in the form of governance networks. Governance networks, in this approach, are vehicles of regulated self-regulation. Governance networks regulate themselves, but there is supervision by governmental organizations that position themselves in the background of the problem solving process. Legitimacy can also be an important reason to create a governance network to solve societal problems. Transferring the solving of societal problems from

(hierarchical) government to a governance network can bring more legitimacy, since the involved actors can play a larger role in the process of problem solving (Provan and Kenis, 2008: 236). Torfing (2012: 104), for example, refers to the importance of governance networks for elected politicians. They form governance networks to gain more legitimacy, since output from these governance networks comes from a bottom-up process and has gained acceptance throughout the involved parties.

(16)

The governance networks in the four Dutch regions in this research where

established because of a combination of the reasons that are given above. Much has already been said about the interdependency of municipalities, housing associations and labour associations in solving the problems surrounding the housing of EU migrant workers. This makes an obvious reason to build up regional governance networks. Furthermore, different resources are needed to solve these problems, which can only be accumulated through active collaboration between the involved actors. The functionalist approach, which argues that actors within the domains of the market and hierarchy should work together to solve complex problems, is also

important for the cases in this research. A combination of hierarchy (governmental organizations, such as municipalities) and markets (housing associations and labour associations) is needed in order to tackle the problem of housing EU migrant workers that finds itself in both the domains of hierarchy and markets.

3.3. What are the specific characteristics of a successful governance network?

Most literature on governance networks is descriptive of nature (Provan and Kenis, 2008). This literature mostly focuses on what governance networks are and on how they come about, which is elaborated on earlier in this chapter. Several researchers (e.g. Klijn and Koppenjan, 2012; Torfing, 2012; Torfing and Sorensen, 2014) in this field of research have described the emerging concept of governance networks and the many reasons why governance networks would be more effective in tackling societal problems than more hierarchical forms of government.

As Provan and Kenis (2008) state in their research, a gap does exist in the literature on governance networks when it comes to measuring the success of these

governance networks. Provan and Kenis (2008) try to close this gap by formulating a few key aspects that are of importance when evaluating the strength of a governance network. These are characteristics of governance networks that can be coupled to successful governance networks. In the introduction of this research, the key characteristics of successful governance networks were already mentioned. These characteristics are: Deemed necessity, goal consensus, active actors in the network, trust and steering. These governance network characteristics will now be discussed.

3.3.1. Network characteristic: Deemed necessity

The first network characteristic of a successful governance network is deemed necessity. This characteristic indicates the opinion of the actors within the network on the necessity of the network for problem solving. It is the need for network-based competencies that is of importance for this characteristic (Provan and Kenis, 2008: 240), the need for a governance network to overcome the problem(s) that the

different actors involved are facing. There are many reasons why a network solution is thought to be the best option to solve certain problems. Provan and Kenis (2008: 240) elaborate on some of these reasons in their research: “the need to gain

legitimacy, serve clients more effectively, attract more resources, addressing complex problems” (Provan and Kenis, 2008: 240). For an effective governance

(17)

network, aimed at a proper output (in the sense of reaching solutions for the problems that are faced by the actors within the network), a majority of the actors within the network should see the relevance of forming a governance network. Furthermore, these actors should acknowledge the positioning of a governance network solution above other (more hierarchical) forms of problem solving.

3.3.2. Network characteristic: Goal consensus

Another important network characteristic of a successful governance network is goal consensus. Goal consensus can have a major influence on the effectiveness of governance networks. For the governance network to be most successful, all involved actors should acknowledge the joint goal that they have to strive for. The argument here is that when there is a common goal, fewer conflicts will occur, which will have a positive influence on the overall effectiveness of the network (Provan and Kenis, 2008: 239). According to Van de Ven (1976: 31) goal consensus among the different actors in the network has a large positive influence on how successful the network is. Besides these positive notes on goal consensus, it should be stated, as Provan and Kenis (2008: 239) do, that the absence of conflicts is not always a good thing, since conflict can have a positive influence by stimulating innovation in the network. Goal consensus is thus of particular interest for the success of a

governance network of any form, but it is not the case that networks with a low degree of goal consensus are not effective at all. We have to keep in mind that when disagreement about the joint goal exists, the other network characteristics together can nullify this disagreement about the goal, and a positive output can come forth from the governance network. However, the most successful are governance

networks in which there is goal consensus among the different actors in the network.

3.3.3. Network characteristic: Active actors in the network

The degree of activity of actors in a governance network has its influence on the success of the governance network. Provan and Kenis (2008:10) argue that the activity of participants influences the success of the governance network. Obviously, when only some actors are active within the network, the network will probably have a lack of proper output. Moreover, the inactivity of some groups of actors can

negatively influence the activity of other actors. It is therefore important that there is a proper division in the amount of activeness of the different actors involved in the governance network.

3.3.4. Network characteristic: Trust

Trust is an important network characteristic of a successful governance networks. Mcevily, Perrone, and Zaheer (2003: 92) describe trust as: “the willingness to accept vulnerability based on positive expectations about another’s intentions or behaviours” (Mcevily, Perrone and Zaheer, 2003: 92). Provan and Kenis (2008: 10) state that the distribution of trust among the participations in network governance is of specific importance for governance network success. Success is more likely when trust is distributed among all participants in the network, which prevents the forming of separate cliques that distrust others outside of their clique. Additionally, Ansell and

(18)

Gash (2008) mention trust as one of the starting conditions for successful network governance. There must be a high density of trust among the different actors within the network (Provan and Kenis, 2008: 238), which means that there are trust-links distributed between the different actors in the network and that “a dense web of trust-based ties” (Provan and Kenis, 2008: 238) is formed. According to a study by

Edelenbos and Klijn (2007: 30), in which they asked for the most important condition for successful partnership, more than 53 per cent of the respondents referred to trust as the most important condition. Trust is therefore probably the most important network characteristic when evaluating the success of a governance network.

3.3.5. Network characteristic: Steering

The previously explained network characteristics (deemed necessity, goal

consensus, active actors in the network and trust) are all internal characteristics of a successful governance network. The next characteristic of governance networks, steering, differs from them in the sense that it can be internal (steering can be done by actors within the network), as well as external (it can be done by actors outside of the network). Some actors within the network are able to play a steering role. There are many definitions of steering, but in this research, the definition by Crawford (2006: 453) is used: Steering is “governing by setting the course, monitoring the direction and correcting deviations from the course set” (Crawford, 2006: 453). In a governance network, this happens in the context of societal problem solving.

Steering can be a vital necessity within the network, since it can affect the output and outcome of the network governance. Most often, governmental actors are interested in steering the network; these actors are advocates of the public interest. However, steering can also be done by non-governmental (private) actors. Klijn, Koppenjan and Termeer (1995: 442) mark that obviously the actor who steers the governance network does not have to be without any interest in steering the network. By steering the network, this actor is able to influence the outcome and thereby to steer the network in the direction compatible with the most favourable outcome for itself. There are many ways in which an actor can steer the governance network, namely: (non-)activation of actors, mobilization of resources, by using interaction rules, and by managing perceptions. The steering actor, when composing the governance network, is able to engage actors within the network that are on the same page as the steering actor itself. Following a strategy called ‘selective activation’, the steering actor can block or activate certain actors from the network and forge the best

coalition for the outcome that the steering actor has in mind (Klijn, Koppenjan and Termeer (1995: 444)). The steering actor has to decide which actors are suited to join the governance network. This actor also has to keep in mind that it is desirable to keep certain actors out of the network, since they can hamper the policy making (Klijn, Koppenjan and Termeer, 1995: 444). Furthermore, resources are an important characteristic that actors within a governance network can have. By forging the network, the steering actor has to keep in mind, which resources the actors possess and which resources are necessary for reaching the certain policy. In this manner,

(19)

resources are an important consideration for the steering actor when selecting which actors are wished for to join the network (Klijn, Koppenjan and Termeer (1995: 445). Rules are also an important characteristic of governance networks. Rules organize a network and can be used consciously or unconsciously (Klijn, Koppenjan and

Termeer, 1995: 444). When the network has already existed for a long period, the steering actor has to bear the rules of the network in mind, since they are able to block or enable the steering (Klijn, Koppenjan and Termeer, 1995: 445). When the network is relatively new, there may be no established rules. In this case, the

steering actor is relatively free and can set the rules by itself. Perceptions are mainly held by all actors within the network (Klijn, Koppenjan and Termeer, 1995: 444). Perceptions when existing can block a certain policy. The steering actor has to try actively to change the different perceptions the actors within the network have and thereby forge the ‘winning coalition’. Redefinition of problems and package deals, in which outcomes are tied to each other, are examples of changing the perceptions of the actors within the network.

3.3.5.1.Tensions in steering the network

Besides these different possibilities available to the steering actor in the governance network, Provan and Kenis (2008: 242-246) refer to three tensions that the steering actor has to deal with, the tensions of efficiency versus inclusiveness, internal versus external legitimacy, and flexibility versus stability. It must furthermore be noted that these tension can also apply to other non-steering actors in the network.

Including many actors within the governance network increases the internal

legitimacy of the network (Provan and Kenis, 2008: 242). More actors involved in the governance network means more influence of different actors and the building of trust. On the other hand, an increase in actors goes together with a decrease in effectiveness of the network, more actors means a more complex network and complexity hampers effectiveness. Furthermore, there is a tension between internal and external legitimacy for governance networks (Provan and Kenis, 2008: 243). The internal legitimacy has to be present in order to make the network effective. Different actors have to be on the same page and a certain degree of trust has to be present. At the same time, there has to be external legitimacy, which means that the network has to be seen legitimate by actors outside of the network. According to Provan and Kenis: “The tension arises when the internal legitimacy needs of network members conflict with external demands” (Provan and Kenis, 2008: 244). There is a constant conflict for actors who have to accomplish their own interests and at the same time have to accomplish the common goals of the network. The last tension is the tension between flexibility and stability (Provan and Kenis, 2008: 244). The strength of

networks is their flexibility. The flexibility of governance networks is one of their advantages over more hierarchical forms of government. On the other hand, when a network becomes more institutionalized over time, more standard procedures etc. will be established and stability will be reached. This stability can hamper the flexibility. These tensions have to be monitored by the steering actor of the network. The

(20)

steering actor, when present, should balance between the tensions in order to make the governance network as effective as possible when it comes to output.

3.4. Conceptual model of a successful governance network

In this part of the theoretical chapter, the conceptual model of a successful

governance network will be drawn and explained. This conceptual model contains the different variables used in this research and the context factor refugee crisis. With this conceptual model, it is possible to image the variables and the links between these variables in one brief overview. This conceptual model is the most wishable situation in which a governance network can exist, when looking at its success. It is expected that when the four regional governance networks in this research have the same network characteristics as given in this model, there will compliance with the governance networks agreements on the housing of EU migrant workers.

The outside box in which the conceptual model is drawn is the contextual factor, the refugee crisis. This is a contextual factor that is expected to influence the success of the governance networks in this research. Then there are the four governance network characteristics that influence each other: deemed necessity, goal

consensus, active actors in the network and trust. These network characteristics are expected to be present in a successful governance network. The governance

network characteristics are influenced by the independent variable steering, since this variable has its influence directly on the dependent variable (the compliance with the agreements on the housing of EU migrant workers), as well as indirectly via the

(21)

governance network characteristics. Therefore, steering is not included in the box with the four other governance network characteristics.

(22)

4. Methodology

This research will compare four regional governance networks in the Netherlands, the governance networks of Holland Rijnland, West-Friesland, Metropoolregio Eindhoven, and Noord-Veluwe. These governance networks are examined in terms of their compliance with the agreements on the housing of EU migrant workers signed in the four regions. Compliance with the regional agreements is seen as governance network success in this research. In this chapter, the methodological framework of this research will be build and discussed. This chapter will be divided into different parts. Firstly, there will be an explanation of the type of research conducted. Then, the case selection will be elaborated on and justified. Next, the coding process will be explained. Thereafter, the central concepts and variables of this research will be operationalized and a conceptual model of a successful governance network will be given. Lastly, the validity and reliability of this research will be discussed.

4.1 Type of research

This research is a small qualitative, comparative case study. There is not one clear definition on what a case study specifically entails. Yin (1994) points to the fact that it is a type of qualitative research and that it has a small N. The latter means that there is only one research object or a small number of research objects. An additional characteristic is that the research takes place “in the field” (Yin, 1994), which entails participant observation. In this study participant observation is seen in the method of gathering data; namely by interviewing the participants in the cases (in the

governance networks) themselves. This method of research has been chosen for several reasons. Gerring (2004: 352) states that one must choose a case study type of research when, besides other reasons, it is important to:

a) Get in-depth knowledge on the relationships between the independent variables and the dependent variable.

b) The internal differentiation between the selected cases is deemed more important than the external generalization of the outcomes.

c) When the examination of the cases is more of an explanatory nature than of a hypotheses testing nature.

All these conditions apply to this research. The data for this research is gathered by conducting a series of qualitative interviews. Most of the interviews were conducted face-to-face; the others were conducted in a written manner. Which interviews were conducted face-to-face and which interviews in a written manner can be found in table 1. By conducting these interviews, insight was gathered in the working of the four regional governance networks. The interviews were conducted in order to measure the five network characteristics for a successful governance network. Governance network success is in the research seen as compliance of actors within

(23)

the governance network with the agreement on the housing of EU migrant workers. A topic list was composed and used, instead of fixed questions, to make adapting to the answers of respondents easier and therefore to conducted a more fluent and open interview.

4.2 Case selection

The cases in this research were not chosen at random, because by choosing cases randomly in a small N case study serious problems are likely to develop (Seawright and Gerring, 2008). When selecting the cases for a small N comparative case study, one has to keep in mind that it requires “1) a representative sample and 2) useful variation on the dimensions of theoretical interest” (Seawright and Gerring, 2008: 296). The particular cases in this research were selected because they were the most similar to each other. According to Seawright and Gerring (2008), selecting cases on most similarities results in outcomes that highlight the differences between cases. In this manner, it will be able to compare some cases in terms of specific variables (in this research these variables are the network characteristics). When the Dutch national government appointed the nine priority regions in The Netherlands, all regions were allowed to come up with their own way of reaching the goal of realizing more housing for EU migrant workers. The four cases in this research (Holland Rijnland, West-Friesland, Metropoolregio Eindhoven and Noord-Veluwe) all took a similar approach towards the task. In all four regions, a non-binding agreement was signed by municipalities, labour associations and housing associations in order to form a governance network that attempts to reach the goal of realizing more suitable housing for EU migrant workers. By selecting these four regions, a most-similar case design could be established. In this way, it is possible to 1) gather in-depth

knowledge on the working of governance networks and 2) highlight the differences between the cases in this most-similar case study design. By gathering this

knowledge, a comparison can be made between the four regional governance networks their approach towards the housing of EU migrant workers. Differences between these similar cases will be highlighted in this comparison. This most-similar case approach is used in this research in order to highlight the differences between the cases when it comes to the governance network characteristics. By researching these differences, knowledge is gathered about which network characteristics account for difference on the dependent variable. By doing so, the causes of differences in the dependent variable, that measures the compliance of the four regional governance networks with the housing agreements, can be verified. In the following section, the four cases in this research, as well as the respondents that were interviewed in these regions, will be presented.

4.2.1 Holland Rijnland

Holland Rijnland is a region in the province of Zuid-Holland. The region is centred around the Dutch city of Leiden and exists of fourteen municipalities. Around 550.000 people live in this region. Four interviews were conducted in the region Holland Rijnland. The municipalities’ view is represented in the findings from the interviews with the municipality of Noordwijkerhout and the municipality of Katwijk. The

(24)

respondent from the municipality of Katwijk also has duties towards the entire region Holland Rijnland and has a place in its ambassador’s team (for more information on the ambassador’s team, see paragraph 3.4.4). The labour associations’ view will be represented by the findings of the interview with the labour association Flexible Human Resources. The respondent from this labour association is also part of the ambassador’s team of the region Holland Rijnland. The housing associations’ view in the region of Holland Rijnland will be represented by the interview with the housing association De Vooruitgang. The respondent from this housing association is also part of the regional ambassador’s team. By interviewing these four respondents, a general overview of the housing of EU migrant workers in the region Holland Rijnland can be given.

4.2.2 West-Friesland

The region West-Friesland is a region in the province Noord-Holland. The region is situated north of Amsterdam and counts twelve municipalities. Around 385.000 people live in this region. Four interviews were conducted in West-Friesland. The findings from the interview with the municipality of Medemblik represents the broader municipalities’ view. The respondent from this municipality is also part of the

ambassador’s team of the region of West-Friesland. The labour associations’ view will be represented by the findings from the interview with labour association AB Vakwerk. The housing associations’ view will be made clear from the findings of two interviews. The interview with the housing association Woningstichting Het

Grootslag, whereof the respondent is also part of the ambassador’s team of the region West-Friesland. As well as the findings of the interview with Rotteveel M4; a private housing association. By interviewing these four respondents, a general overview of the housing of EU migrant workers in the region West-Friesland can be established.

4.2.3 Metropoolregio Eindhoven

The region Metropoolregio Eindhoven is a region in the province Noord-Brabant. The city of Eindhoven lies at the heart of this region, which counts twenty-one

municipalities. Around 750.000 people live in this region. In the region Metropoolregio Eindhoven three interviews were conducted. The municipalities’ view is represented from the findings of the interview with the municipality of Helmond. The respondent from this municipality is also a member of the ambassador’s team in his region. The labour associations’ view will be found in the interview with the labour association Z.O.N. Arbeidsvoorziening. The view of the housing associations in the

Metropoolregio Eindhoven is expressed by the interview with the housing association Thuis in Eindhoven. By conducting these three interviews, a general picture of the housing of EU migrant workers in the Metropoolregio Eindhoven is brought into view.

4.2.4 Noord-Veluwe

The region Noord-Veluwe is a region in the province Gelderland. The region is centred around the city of Harderwijk and consists of ten municipalities. Around 250.000 people live in the region. In the region Noord-Veluwe, four interviews were

(25)

conducted. The municipalities’ view is represented by the findings of the interview with a respondent from the regional government of Noord-Veluwe, as well as those from the interview with the municipality of Harderwijk. The respondent from the municipality of Harderwijk is also a member of the regional ambassador’s team. The housing associations’ view is represented by the interview with housing association Omnia Wonen. Unfortunately it was, despite a lot of effort, not possible to interview a labour association from this region, since none of the labour associations in the region was willing to be interviewed about the subject. The lack of the labour

associations’ view in the region Noord-Veluwe will be partly remedied by conducting the other three interviews. By conducting these interviews, a general picture is given of the housing of EU migrant workers in the region Noord-Veluwe.

In Table 1, the different organizations that were interviewed can be found.

Table 1: Respondents in the four regional governance networks

Region Organization Member of the

Ambassador’s Team

Face-to-face or written interview Holland Rijnland Municipality:

Gemeente Noordwijkerhout

No Face-to-face

Municipality: Gemeente Katwijk and region Holland Rijnland Yes Face-to-face Labour association: Flexible Human Resources Yes Face-to-face Housing association: De Vooruitgang Yes Written West-Friesland Municipality: Gemeente Medemblik Yes Face-to-face Labour association: AB Vakwerk No Face-to-face Housing association:

Het Grootslag Yes Face-to-face

Private housing association: Rotteveel M4 No Face-to-face Metropoolregio Eindhoven Municipality: Gemeente Helmond Yes Face-to-face Labour association: Z.O.N. Arbeidsvoorziening No Written Housing association: Thuis Yes Face-to-face Noord-Veluwe Municipality: Gemeente Ermelo and region

(26)

Veluwe Municipality: Gemeente Harderwijk Yes Written Housing association: Omnia Wonen No Face-to-face

Table 1: In this table the actors that were interviewed are given. For each of the actors there can be found if this actor is part of the regional ambassador’s team and if it was a face-to-face interview or a written interview.

4.3 Coding

As a method of analysing the qualitative data that was gained by interviewing the respondents for this research, the data was coded. In the coding process, the coding software Atlas.ti (version 7) was used. The codes that were used in this research were deductively generated from the different network characteristics for governance network success that were used in this research. These codes correspond with the indicators that measure the different governance network characteristics for

governance network success. Furthermore, these codes were grouped in code families. The code families were named after the network characteristics of the research. In this manner, all codes that said something about for example the network characteristic trust in a governance network were placed in the

corresponding code family. For each network characteristic there were two code families. One code family with the label ‘positive’ and one code family with the label ‘negative’. This in order to say something about a specific network characteristic for governance network success in a region. In this way, it was possible to say

something about, for example the trust, in one of the four regional governance networks. The labels positive or negative were not placed on the codes randomly. A positive label was added to a code when there was positively referred to one of the indicators by the respondent. A negative label was added to a code when there was negatively referred to one of the indicators by the respondent. In this way, it was possible to count the different positive and negative codes that were placed on the text in order to make an image of the different regional governance networks for each of the network characteristics. When there was an extensive majority of either

positive or negative codes for a specific network characteristic, the network characteristic was given this outcome for the specific region. When there was no extensive majority for either of the labels, the outcome mixed was given to this network characteristic in a region. In this way, it was possible to compare the four regional governance networks for each of the network characteristics for governance network success.

4.4 Description of central concepts

Some concepts have a vital role in this study. In order to conduct proper research, these concepts must be consistently described and used. In this section of the thesis, the most important concepts will be described in the following order: EU migrant

(27)

workers, governance network, governance network branche, housing agreement and ambassador’s team.

4.4.1 EU migrant workers

In this study, the term EU migrant worker refers to:

“EU migrant workers are people between 15 and 64 years old,

originated from the countries: Bulgaria, Hungary, Slovenia, Poland, Rumania, Estonia, Latvia, Lithuania, Czech Republic, Slovakia, Portugal, Spain, Italy and Greece that are from the first generation and are living in The Netherlands on the 31th December 2012 ”

(Quick Scan Holland Rijnland, 2013).

This definition is focussed on the Netherlands, since all regions in this research are situated within The Netherlands. Furthermore is chosen for this definition because this definition is also used as a definition of EU migrant workers in policy documents from the regions in this research. One can separate EU migrant workers according to their length of stay in three groups: short-stay, mid-stay and long-stay. More about these forms of stay and the effects of this division for the policy on the housing of EU migrant workers is discussed in paragraph 2.2.

4.4.2 Governance network

In this research, the definition of a governance network as given by Torfing (2012) is used:

“Horizontal articulations of interdependent, but operationally

autonomous, actors from the public and/or private sector who interact with one another through ongoing negotiations that take place within a regulative, normative, cognitive, and imaginary framework; facilitate self-regulation in the shadow of hierarchy; and contribute to the production of public regulation in the broad sense of the term”. (Torfing, 2012: 101)

The governance networks in the four regions in this research all fit the mould of Torfing (2012) his definition. The actors in the governance networks are

municipalities, labour associations and housing associations.

4.4.3 Governance network branche

In the four regional governance networks the actors can broadly be divided into three branches. These branches correspond with the backgrounds or sectors that these actors have. The three branches are: Municipalities, housing associations and labour associations. In the branche municipalities, all municipalities that were interviewed are found. The housing association branche consists of actors who have the core task of realizing housing. The labour association branche consists of actors which are the employers of the EU migrant workers.

(28)

4.4.4 Housing agreement

In all four regions in this research, a non-binding agreement was signed. This

agreement established a governance network to realize more and better housing for EU migrant workers. These non-binding agreements between the involved actors in the governance networks are referred to as housing agreements.

4.4.5 Ambassador’s team

In each of the four regions in this research an ambassador’s team was formed. The main task of this ambassador’s team was to bring the three most important sectors in the housing of EU migrant workers (municipalities, labour associations and housing associations) together. To make this task easier, the ambassador’s team in all four regions consisted of at least one representative from each of the three main sectors. Bringing the different sectors together was mainly done by being the driving force in the process of problem solving.

4.5 Operationalization of variables

In this part of the research, the dependent and independent variables will be operationalized. For all variables, at least two indicators were deducted from the theories on network characteristics described in the theoretical chapter of this

research. In this way, it was possible to measure each of the network characteristics for all four regions and therefore to say something about the links between the network characteristics and governance network success. In table 2, the variables (network characteristics) of this research are listed, accompanied by a theoretical definition, an operational definition and the indicators. As described in the paragraph on the data coding all codes were, corresponding with the indicators of this research, grouped according to a positive or negative load. By counting the number of positive and negative codes for each of the variables, it became possible to map the four regional governance networks for the determents of governance network success.

Table 2: Operationalization of variables

Variable Theoretical definition Operational definition Indicators of the variable Compliance with the governance network's agreement on the housing of EU migrant workers Compliance is the degree in which the governance network is successful in realizing more housing for EU migrant workers Compliance measures the number of beds realized by the governance network in each of the four regions, according to a governmental document from the first of January 2016; Huisvesting

EU-arbeidsmigranten: Stand van zaken in de regio’s per januari 2016 Percentage consisting of the absolut number of realized beds in each region divided by the goal set in each of the four regions for the number of beds that had to be realized

(29)

(Rijksoverheid, 2016).

Deemed necessity “Deemed necessity indicates the opinion of the actors within the network on the necessity of the network for problem solving” (Provan and Kenis, 2008: 240)

Deemed necessity measures the need for cooperation for solving the housing problem and the acknowledgement of a governance network solution as best option for solving the housing problem Need for cooperation Acknowledgement of a governance network solution as best option for solving the problems

Goal consensus Goal consensus is the degree of consensus among the actors in the governance network about the joint goal that the network should pursue

Goal consensus measures the acknowledgement of the majority of the actors in the network of the housing problem as a problem that needs to be solved and the goal consensus between the three main sectors in a region (municipalities, labour associations and housing associations) Acknowledging of the majority of the actors within the network of the housing problem as a problem that needs to be solved Goal consensus between the three branches (cliques) in a region (municipalities, labour associations and housing associations) Active actors in the

network

The degree of activity of actors within the network

Active actors in the network measures the degree of initiatives of the different branches for realizing more housing for EU migrant workers and the willingness of each of these branches in

cooperating with the other branches in solving the problem

Initiatives of the different branches for realizing more housing for EU migrant workers Willingness of a branch to cooperate with the other branches in solving the housing problem

Trust “the willingness to accept vulnerability based on positive expectations about another’s intentions or behaviours” (Mcevily, Perrone and Zaheer, 2003: 92)

Trust measures the judgement about the efforts of other actors and the trust of the three main sectors in a region (municipalities, labour associations and housing

associations) in each other in the

Judgement about the efforts of other actors Trust of different branches in a region (cliques) in each other (municipalities, labour association and housing

Network governance. A comparison of four Dutch regions in their approac towards the housing of EU migrant workers